Why Europe is Building Empty Data Centres and Calling It Strategy

Why Europe is Building Empty Data Centres and Calling It Strategy

The industry consensus is a comfort blanket woven from denial. Read any mainstream trade publication right now, and you will find analysts swooning over European legacy carriers pivoting into real estate developers, throwing billions at data centres and server sheds to capture the artificial intelligence boom. The narrative says heritage operators like Orange, Vodafone, and Deutsche Telekom are future-proofing their balance sheets by transforming from dumb pipes into digital landlords.

It is a fairy tale told by people who have never had to slash operating expenditure while drowning in debt.

I have watched legacy operators blow millions on infrastructure pivots that treat symptoms while ignoring structural insolvency. Transforming a telecommunications provider into a data centre operator does not fix a broken core business model; it merely slaps fresh coats of paint on a sinking ship. The math driving this supposed craze does not survive contact with reality, and the strategic rationale collapses the moment you examine Europe's structural power constraints, capital expenditure realities, and regulatory chokeholds.

The Capital Expenditure Trap

Let us look at the fundamental economics. Building modern hyperscale compute infrastructure requires massive, front-loaded capital outlays with agonizingly long payback periods. European telcos are not Silicon Valley hyperscalers printing free cash flow from global software monopolies. They are mature utilities carrying heavy debt loads, burdened by legacy pension obligations, and locked into hyper-competitive consumer markets where ARPU (Average Revenue Per User) is among the lowest in the developed world.

When a carrier diverts billions away from network modernization—such as rolling out dense fibre and standalone 5G—to fund concrete bunkers full of GPUs, they are robbing Peter to pay a landlord who hasn't secured tenants.

Imagine a scenario where an indebted carrier spends three billion euros on a state-of-the-art facility in Frankfurt or Madrid, only to find that enterprise customers prefer to deploy workloads directly with American hyperscalers who own the proprietary silicon stack and developer ecosystems.

You do not win the compute market by building a shed and hoping someone rents a rack. You win by owning the software stack, the energy supply, and the chip design. European telcos own none of the three.

The Energy Illusion

The lazy consensus assumes that because telcos already own prime real estate across major European cities and transport corridors, converting local exchanges and central offices into edge data centres is a natural synergy.

Physics disagrees.

Data centres do not run on corporate optimism; they run on uninterrupted, mega-watt scale electricity. Europe’s power grid is fragmented, heavily constrained, and burdened by some of the highest industrial energy tariffs on the planet. While American tech giants can anchor their massive data investments near cheap geothermal, hydro, or nuclear power, European operators are tethered to a grid struggling under the weight of de-industrialization and regulatory green targets.

If you pack high-density AI accelerators into repurposed urban telephone exchanges designed for copper-line switches from the 1980s, you will melt your cooling systems before you compute a single token. Upgrading the local grid connection for high-density compute requires permitting processes that take years, completely neutralizing any supposed speed-to-market advantage.

Sovereignty as a Slogan

Proponents of the telco data centre pivot love to wave the banner of European digital sovereignty. The argument goes that local enterprises need sovereign cloud environments hosted by trusted regional telecom brands to protect data from extraterrestrial regulatory reach.

This is a boardroom delusion designed to appeal to Brussels regulators.

Sovereignty without competitive performance is just expensive isolationism. Enterprise customers—whether German automotive manufacturers or French pharmaceutical giants—do not choose their cloud provider based on patriotic sentiment. They choose based on latency, developer tooling, model efficiency, and total cost of ownership. When a domestic carrier offers a second-rate cloud managed through bureaucratic joint ventures, enterprises route around them and use global hyperscalers anyway, compliance risks be damned.

We saw this exact dynamic play out with previous enterprise cloud joint ventures launched by European carriers. They were heralded with massive PR campaigns, heavily praised in policy whitepapers, and quietly downsized within three years because they could not out-innovate or out-price the American and Asian cloud oligopolies.

The Real Play They Are Missing

If European operators want to survive, they need to stop trying to out-build companies whose core competency is compute infrastructure. They should lean into what they actually own: the last-mile network edge and local subscriber trust, and monetize it through intelligent API exposure rather than pouring concrete.

The data centre craze is a desperate distraction. It is a shiny object for executives who lack the strategic imagination to fix stagnant core connectivity revenues. Stop praising the pivot. It is capital destruction disguised as digital transformation.

EP

Elena Parker

Elena Parker is a prolific writer and researcher with expertise in digital media, emerging technologies, and social trends shaping the modern world.